In [1]:
# importing libraries
import datetime
import os
import sys
import random
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
import warnings
warnings.filterwarnings('ignore')
from plotly import tools, subplots
import plotly.offline as py
from plotly.offline import plot
py.init_notebook_mode(connected = True)
import plotly.graph_objs as go
import plotly.express as px
import plotly.figure_factory as ff
import plotly.io as pio
from datetime import datetime
In [2]:
# Loading Data
confirmed_global = pd.read_csv('time_series_covid19_confirmed_global.csv')
deaths_global = pd.read_csv('time_series_covid19_deaths_global.csv')
recovered_global = pd.read_csv('time_series_covid19_recovered_global.csv')
confirmed_us = pd.read_csv('time_series_covid19_deaths_US.csv')
deaths_us = pd.read_csv('time_series_covid19_deaths_US.csv')
In [ ]:
In [3]:
confirmed_global
Out[3]:
In [4]:
# Renaming the columns
confirmed_global.rename(columns = {
'Country/Region':'Country_Region',
'Province/State': 'Province_State',
'Longitude': 'Long',
'Latitude': 'Lat'
}, inplace = True)
recovered_global.rename(columns = {
'Country/Region':'Country_Region',
'Province/State': 'Province_State',
'Longitude': 'Long',
'Latitude': 'Lat'
}, inplace = True)
deaths_global.rename(columns = {
'Country/Region':'Country_Region',
'Province/State': 'Province_State',
'Longitude': 'Long',
'Latitude': 'Lat'
}, inplace = True)
In [5]:
#calculatinng the NaN values
print(confirmed_global.isnull().sum())
print(recovered_global.isnull().sum())
print(deaths_global.isnull().sum())
In [6]:
# replacing Nan values to string values 'nan'
confirmed_global["Province_State"].fillna("nan", inplace = True)
recovered_global["Province_State"].fillna("nan", inplace = True)
deaths_global["Province_State"].fillna("nan", inplace = True)
In [7]:
confirmed_global.Country_Region.unique()
Out[7]:
In [8]:
confirmed_global = confirmed_global[~confirmed_global["Province_State"].str.match('Diamond Princess')]
confirmed_global = confirmed_global[~confirmed_global["Country_Region"].str.match('Diamond Princess')]
deaths_global = deaths_global[~deaths_global["Province_State"].str.match('Diamond Princess')]
deaths_global = deaths_global[~deaths_global["Country_Region"].str.match('Diamond Princess')]
recovered_global = recovered_global[~recovered_global["Province_State"].str.match('Diamond Princess')]
recovered_global = recovered_global[~recovered_global["Country_Region"].str.match('Diamond Princess')]
In [9]:
# declaring function for converting Date formats
def convert_date(data):
try:
data.columns = list(data.columns[:4]) + [datetime.strptime(dt, "%m/%d/%y").date().strftime("%Y-%m-%d") for dt in data.columns[4:]]
except:
data.columns = list(data.columns[:4]) + [datetime.strptime(dt, "%m/%d/%Y").date().strftime("%Y-%m-%d") for dt in data.columns[4:]]
In [10]:
# calling function to change date formats
convert_date(confirmed_global)
convert_date(recovered_global)
convert_date(deaths_global)
In [11]:
confirmed_global
Out[11]:
In [12]:
confirmed_global_df = confirmed_global.melt(id_vars = ['Country_Region','Province_State','Lat','Long'],
value_vars = confirmed_global.columns[4:],
var_name = 'Date',
value_name = 'Confirmed_Cases')
In [13]:
deaths_global_df = deaths_global.melt(id_vars = ['Country_Region','Province_State','Lat','Long'],
value_vars = confirmed_global.columns[4:],
var_name = 'Date',
value_name = 'Deaths')
In [14]:
recovered_global_df = recovered_global.melt(id_vars = ['Country_Region','Province_State','Lat','Long'],
value_vars = confirmed_global.columns[4:],
var_name = 'Date',
value_name = 'Recovered')
In [15]:
recovered_global_df
Out[15]:
In [16]:
train = confirmed_global_df.merge(deaths_global_df, on = ['Country_Region', 'Province_State','Date'])
train = train.merge(recovered_global_df, on = ['Country_Region', 'Province_State','Date'])
In [17]:
global_dataset = train.groupby('Date')['Confirmed_Cases','Recovered','Deaths'].sum().reset_index()
In [18]:
global_dataset['Daily_Rise'] = global_dataset['Confirmed_Cases'] - global_dataset['Confirmed_Cases'].shift(1)
global_dataset['Mortality_Rate'] = global_dataset['Deaths']/ global_dataset['Confirmed_Cases']
global_dataset
Out[18]:
In [19]:
global_dataset_df = pd.melt(global_dataset,
id_vars = ['Date'],
value_vars = ['Confirmed_Cases','Recovered', 'Mortality_Rate', 'Deaths', 'Daily_Rise'])
global_dataset_df
Out[19]:
In [20]:
visual_confirmed = global_dataset_df[global_dataset_df["variable"].str.match('Confirmed_Cases')]
visual_deaths = global_dataset_df[global_dataset_df["variable"].str.match('Deaths')]
visual_recovered = global_dataset_df[global_dataset_df["variable"].str.match('Recovered')]
visual_mortality = global_dataset_df[global_dataset_df["variable"].str.match('Mortality_Rate')]
visual_daily_rise = global_dataset_df[global_dataset_df["variable"].str.match('Daily_Rise')]
In [21]:
visual_confirmed
Out[21]:
In [22]:
visual_deaths
Out[22]:
In [23]:
visual_recovered
Out[23]:
In [24]:
visual_mortality
Out[24]:
In [25]:
visual_daily_rise
Out[25]:
In [26]:
fig = px.line(global_dataset_df,
x = 'Date',
y = 'value',
color = 'variable',
title = 'Global Confirmed/ Deaths/ REcovered/ cases with Mortality and Daily Rises')
fig.show()
In [27]:
fig = px.line(global_dataset_df,
x = 'Date',
y = 'value',
color = 'variable',
title = 'Global Confirmed/ Deaths/ REcovered/ cases with Mortality and Daily Rises',
log_y = True)
fig.show()
In [28]:
fig = px.line(visual_confirmed,
x = 'Date',
y = 'value',
color = 'variable',
title = 'Confirmed cases over time(Globally)')
fig.show()
In [29]:
fig = px.line(visual_deaths,
x = 'Date',
y = 'value',
color = 'variable',
title = 'Deaths reported over time(Globally)')
fig.show()
In [30]:
fig = px.line(visual_mortality,
x = 'Date',
y = 'value',
color = 'variable',
title = 'Change in Mortality Rate over time')
fig.show()
In [31]:
fig = px.line(visual_daily_rise,
x = 'Date',
y = 'value',
color = 'variable',
title = 'Daily rise of the infected people(Globally)')
fig.show()
In [32]:
country_wise = train.groupby(['Country_Region','Province_State','Date'])['Confirmed_Cases','Recovered','Deaths'].sum().reset_index()
In [33]:
country_wise
Out[33]:
In [34]:
country_wise = pd.melt(country_wise,
id_vars = ['Date','Country_Region','Province_State'],
value_vars = ['Confirmed_Cases','Recovered', 'Deaths'])
In [35]:
country_wise
Out[35]:
In [36]:
country_wise_visual_confirmed = country_wise[country_wise["variable"].str.match('Confirmed_Cases')]
country_wise_visual_deaths = country_wise[country_wise["variable"].str.match('Deaths')]
country_wise_visual_recovered = country_wise[country_wise["variable"].str.match('Recovered')]
In [37]:
country_wise
Out[37]:
In [38]:
fig = px.line(country_wise_visual_confirmed,
x = 'Date',
y = 'value',
color = 'Country_Region',
title = 'Confirmed cases')
fig.show()
In [39]:
fig = px.line(country_wise_visual_deaths,
x = 'Date',
y = 'value',
color = 'Country_Region',
title = 'People that are missed')
fig.show()
In [40]:
fig = px.line(country_wise_visual_recovered,
x = 'Date',
y = 'value',
color = 'Country_Region',
title = 'Recovered cases')
fig.show()
In [41]:
current_date = country_wise_visual_confirmed['Date'][country_wise_visual_confirmed.index[-1]]
In [42]:
# finding out total deaths, confirmed cases and recovered cases
country_wise_total_confirmed = country_wise_visual_confirmed[country_wise_visual_confirmed["Date"].str.match(current_date)]
country_wise_total_deaths = country_wise_visual_deaths[country_wise_visual_deaths["Date"].str.match(current_date)]
country_wise_total_recovered = country_wise_visual_recovered[country_wise_visual_recovered["Date"].str.match(current_date)]
In [43]:
country_wise_total_deaths
Out[43]:
In [44]:
country_wise_total_recovered
Out[44]:
In [45]:
country_wise_total_confirmed
Out[45]:
In [46]:
worst_hit_countries_30 = country_wise_total_confirmed.sort_values(by = 'value', ascending = False).head(30)
plt.figure(figsize = (12, 10))
sns.barplot(data = worst_hit_countries_30, y = 'Country_Region', x = 'value', hue = 'Country_Region', dodge = False)
plt.legend(loc = 'lower right')
plt.xlabel('Total Confirmed reported')
plt.ylabel('Countries')
plt.title('Worst 30 countries hit by navirus(Confirmed)')
plt.show()
In [47]:
worst_hit_countries_30 = country_wise_total_deaths.sort_values(by = 'value', ascending = False).head(30)
plt.figure(figsize = (12, 10))
sns.barplot(data = worst_hit_countries_30, y = 'Country_Region', x = 'value', hue = 'Country_Region', dodge = False)
plt.legend(loc = 'lower right')
plt.xlabel('Most number of people missed')
plt.ylabel('Countries')
plt.title('Worst 30 countries hit by navirus(Deaths)')
plt.show()
In [48]:
recovering_countries_30 = country_wise_total_recovered.sort_values(by = 'value', ascending = False).head(30)
fig = px.bar(recovering_countries_30,
x='value', y='Country_Region', color='Country_Region', barmode='relative',
title=f'Most Recovered', text='value', height=1500, width = 950, orientation='h')
fig.show()
In [49]:
locations = confirmed_global[confirmed_global["Province_State"].str.match('nan')].reset_index()
locations = locations[['Country_Region','Lat','Long']]
locations
Out[49]:
In [50]:
country_wise_total_confirmed = country_wise_total_confirmed.groupby(['Country_Region','Date'])[['value']].sum().reset_index()
country_wise_total_deaths = country_wise_total_deaths.groupby(['Country_Region','Date'])[['value']].sum().reset_index()
country_wise_total_recovered = country_wise_total_recovered.groupby(['Country_Region','Date'])[['value']].sum().reset_index()
In [51]:
total_countries = country_wise_total_confirmed.merge(country_wise_total_deaths, on = ['Country_Region','Date'])
total_countries = total_countries.merge(country_wise_total_recovered, on = ['Country_Region','Date'])
In [52]:
total_countries
Out[52]:
In [53]:
country_wise_total_confirmed
Out[53]:
In [54]:
total_countries.rename(columns = {
'value_x': 'Confirmed_Cases',
'value_y': 'Deaths',
'value' : 'Recovered'}, inplace = True)
#total_countries.drop(['variable_x','variable_y','variable'], axis = 1, inplace = True)
In [55]:
fig = px.choropleth(total_countries,
locations = 'Country_Region',
locationmode='country names',
color = 'Confirmed_Cases',
hover_name = 'Country_Region',
color_continuous_scale="Viridis",
scope = 'world',
hover_data = ['Confirmed_Cases','Deaths','Recovered'],
title='Covid19 cases worldwide')
fig.update_geos(fitbounds="locations", visible=True)
fig.update_layout(margin={"r":0,"t":0,"l":0,"b":0})
fig.show()
In [56]:
fig = px.choropleth(total_countries,
locations = 'Country_Region',
locationmode='country names',
color = 'Confirmed_Cases',
hover_name = 'Country_Region',
color_continuous_scale="Viridis",
scope = 'north america',
hover_data = ['Confirmed_Cases','Deaths','Recovered'],
title='Covid19 cases worldwide')
fig.update_geos(fitbounds="locations", visible=False)
fig.update_layout(margin={"r":0,"t":0,"l":0,"b":0})
fig.show()
In [57]:
country_names = []
for i in (total_countries["Country_Region"]):
country_names.append(i)
In [58]:
country_names.remove('Holy See')
country_names.remove('Kosovo')
country_names.remove('MS Zaandam')
country_names.remove('Timor-Leste')
country_names.remove('US')
country_names.remove('West Bank and Gaza')
country_names.remove("Western Sahara")
In [59]:
country_names = [sub.replace('Burma', 'Myanmar') for sub in country_names]
country_names = [sub.replace('Congo (Brazzaville)', 'Congo') for sub in country_names]
country_names = [sub.replace('Congo (Kinshasa)', 'Democratic Republic of the Congo') for sub in country_names]
country_names = [sub.replace('Cote d\'Ivoire', 'Côte d\'Ivoire') for sub in country_names]
country_names = [sub.replace('Korea, South', 'South Korea') for sub in country_names]
country_names = [sub.replace('Taiwan*', 'Taiwan') for sub in country_names]
In [60]:
from pycountry_convert import country_alpha2_to_continent_code, country_name_to_country_alpha2
continents = {
'NA': 'North America',
'SA': 'South America',
'AS': 'Asia',
'OC': 'Australia',
'AF': 'Africa',
'EU': 'Europe'
}
y = [continents[country_alpha2_to_continent_code(country_name_to_country_alpha2(country))] for country in country_names]
In [61]:
#continent and countries
continents_country = pd.DataFrame(list(zip(country_names, y)),
columns =['Country_Region', 'Continent'])
In [62]:
continents_country['Country_Region'] = continents_country['Country_Region'].replace({'Congo':'Congo (Brazzaville)',
'Democratic Republic of the Congo':'Congo (Kinshasa)',
'Côte d\'Ivoire': 'Cote d\'Ivoire',
'South Korea': 'Korea, South',
'Myanmar': 'Burma',
'Taiwan': 'Taiwan*'})
In [63]:
total_countries_and_cont = total_countries.merge(continents_country, on = ['Country_Region'])
In [64]:
african_countries = total_countries_and_cont[total_countries_and_cont["Continent"].str.match('Africa')].reset_index()
asian_countries = total_countries_and_cont[total_countries_and_cont["Continent"].str.match('Asia')].reset_index()
north_american_countries = total_countries_and_cont[total_countries_and_cont["Continent"].str.match('North America')].reset_index()
european_countries = total_countries_and_cont[total_countries_and_cont["Continent"].str.match('Europe')].reset_index()
australian_countries = total_countries_and_cont[total_countries_and_cont["Continent"].str.match('Australia')].reset_index()
south_american_countries = total_countries_and_cont[total_countries_and_cont["Continent"].str.match('South America')].reset_index()
fig = px.choropleth(african_countries,
locations = 'Country_Region',
locationmode='country names',
color = 'Confirmed_Cases',
hover_name = 'Country_Region',
color_continuous_scale="Viridis",
scope = 'africa',
hover_data = ['Confirmed_Cases','Deaths','Recovered'],
title='Covid19 cases worldwide')
fig.update_geos(fitbounds="locations", visible=False)
fig.update_layout(margin={"r":0,"t":0,"l":0,"b":0})
fig.show()
In [65]:
fig = px.choropleth(asian_countries,
locations = 'Country_Region',
locationmode='country names',
color = 'Confirmed_Cases',
hover_name = 'Country_Region',
color_continuous_scale="Viridis",
scope = 'asia',
hover_data = ['Confirmed_Cases','Deaths','Recovered'],
title='Covid19 cases worldwide')
fig.update_geos(fitbounds="locations", visible=False)
fig.update_layout(margin={"r":0,"t":0,"l":0,"b":0})
fig.show()
In [66]:
fig = px.choropleth(south_american_countries,
locations = 'Country_Region',
locationmode='country names',
color = 'Confirmed_Cases',
hover_name = 'Country_Region',
color_continuous_scale="Viridis",
scope = 'south america',
hover_data = ['Confirmed_Cases','Deaths','Recovered'],
title='Covid19 cases worldwide')
fig.update_geos(fitbounds="locations", visible=False)
fig.update_layout(margin={"r":0,"t":0,"l":0,"b":0})
fig.show()
In [67]:
fig = px.choropleth(australian_countries,
locations = 'Country_Region',
locationmode='country names',
color = 'Confirmed_Cases',
hover_name = 'Country_Region',
color_continuous_scale="Viridis",
hover_data = ['Confirmed_Cases','Deaths','Recovered'],
title='Covid19 cases worldwide')
fig.update_geos(fitbounds="locations", visible=False)
fig.update_layout(margin={"r":0,"t":0,"l":0,"b":0})
fig.show()
In [68]:
fig = px.choropleth(european_countries,
locations = 'Country_Region',
locationmode='country names',
color = 'Confirmed_Cases',
hover_name = 'Country_Region',
color_continuous_scale="Viridis",
scope = 'europe',
hover_data = ['Confirmed_Cases','Deaths','Recovered'],
title='Covid19 cases worldwide')
fig.update_geos(fitbounds="locations", visible=False)
fig.update_layout(margin={"r":0,"t":0,"l":0,"b":0})
fig.show()
In [ ]:
In [69]:
confirmed_us.drop(['UID','iso2','iso3','Admin2','Country_Region','code3','Combined_Key','Population'],axis = 1, inplace = True)
In [70]:
confirmed_us
Out[70]:
In [71]:
deaths_us.drop(['UID','iso2','iso3','Admin2','Country_Region','code3','Combined_Key','Population'], axis = 1, inplace = True)
In [72]:
# declaring function for converting Date formats
def convert_date_us(data):
try:
data.columns = list(data.columns[:4]) + [datetime.strptime(dt, "%m/%d/%y").date().strftime("%Y-%m-%d") for dt in data.columns[4:]]
except:
data.columns = list(data.columns[:4]) + [datetime.strptime(dt, "%m/%d/%Y").date().strftime("%Y-%m-%d") for dt in data.columns[4:]]
In [73]:
convert_date_us(deaths_us)
convert_date_us(confirmed_us)
In [74]:
confirmed_us
Out[74]:
In [75]:
from urllib.request import urlopen
import json
with urlopen('https://raw.githubusercontent.com/plotly/datasets/master/geojson-counties-fips.json') as response:
counties = json.load(response)
In [76]:
confirmed_us_df = confirmed_us.melt(id_vars = ['Province_State','FIPS','Lat','Long_'],
value_vars = confirmed_us.columns[4:],
var_name = 'Date',
value_name = 'Confirmed_Cases')
In [77]:
deaths_us_df = deaths_us.melt(id_vars = ['Province_State','FIPS','Lat','Long_'],
value_vars = deaths_us.columns[4:],
var_name = 'Date',
value_name = 'Confirmed_Cases')
In [78]:
confirmed_us_df = confirmed_us_df.groupby(['Province_State','FIPS','Date'])['Confirmed_Cases'].sum().reset_index()
In [79]:
confirmed_us_df.info()
In [80]:
confirmed_us_df['FIPS'] = confirmed_us_df['FIPS'].astype(int)
In [81]:
x = confirmed_us_df.groupby(['Province_State','FIPS'])['Confirmed_Cases'].sum().reset_index()
x
Out[81]:
In [82]:
fig = px.choropleth(x,
geojson=counties,
locations='FIPS', color='Confirmed_Cases',
color_continuous_scale="Viridis",
range_color=(confirmed_us_df['Confirmed_Cases'].min(), confirmed_us_df['Confirmed_Cases'].max()),
hover_name = 'Province_State',
scope = 'usa')
fig.update_layout(margin={"r":0,"t":0,"l":0,"b":0})
fig.show()